Applied Thermal Engineering 51 (2013) 711e722

Contents lists available at SciVerse ScienceDirect

Applied Thermal Engineering

journal homepage: www.elsevier.com/locate/apthermeng

Review of organic Rankine cycles for internal combustion engine

exhaust waste heat recovery
Charles Sprouse III *, Christopher Depcik
University of Kansas, Department of Mechanical Engineering, 3138 Learned Hall, 1530 W. 15th Street, Lawrence, KS 66045, USA

h i g h l i g h t s

< This review article focuses on engine exhaust waste heat recovery works.
< The organic Rankine cycle is superior for low to medium exergy heat sources.
< Working fluid and expander selection strongly influence efficiency.
< Several authors demonstrate viable systems for vehicle installation.

a r t i c l e i n f o a b s t r a c t

Article history: Escalating fuel prices and future carbon dioxide emission limits are creating a renewed interest in
Received 14 May 2012 methods to increase the thermal efficiency of engines beyond the limit of in-cylinder techniques. One
Accepted 13 October 2012 promising mechanism that accomplishes both objectives is the conversion of engine waste heat to
Available online 22 October 2012
a more useful form of energy, either mechanical or electrical. This paper reviews the history of internal
combustion engine exhaust waste heat recovery focusing on Organic Rankine Cycles since this ther-
Keywords:
modynamic cycle works well with the medium-grade energy of the exhaust. Selection of the cycle
Waste heat recovery
expander and working fluid are the primary focus of the review, since they are regarded as having the
Organic Rankine cycle
Engine emissions
largest impact on system performance. Results demonstrate a potential fuel economy improvement
Fuel consumption around 10% with modern refrigerants and advancements in expander technology.
Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction engines have met these standards by using the Three Way Catalyst
(TWC), while Compression Ignition (CI) engines have been fitted
The Internal Combustion Engine (ICE) has been a primary power with a 4-way aftertreatment system [2]. In both cases, catalytic
source for automobiles, long-haul trucks, locomotives, and ships devices convert harmful emissions to nitrogen gas (N2), carbon
over the past century. Over this time, periods of high fuel costs and dioxide (CO2), and water (H2O). Through modification of injection
concerns about foreign oil dependence have resulted in increasingly timing and rate shaping, CI engines often operate with reduced
complex engine designs to decrease fuel consumption. For example, combustion temperatures to prevent the formation of NOx as its
engine manufacturers have implemented techniques such as conversion in a 4-way aftertreatment system is relatively difficult.
enhanced fuel/air mixing, turbocharging, and variable valve timing SI engines often accomplish a similar tactic by retarding the spark
in order to increase thermal efficiency. However, around 60e70% of timing; however, TWC devices have the inherent advantage of
the fuel energy is still lost as waste heat through the coolant or the using CO and HC to reduce NOx. Therefore, there is less of a need to
exhaust [1]. Moreover, increasingly stringent emissions regulations adjust the combustion timing. As mentioned previously, these
are causing engine manufacturers to limit combustion tempera- strategies reduce engine efficiency by achieving less combustion at
tures and pressures lowering potential efficiency gains. prime crank angles.
To date, legislation of exhaust emission levels has focused on It appears that future regulations will focus more on CO2
carbon monoxide (CO), hydrocarbons (HC), nitrogen oxides (NOx), emissions, which require increasing the efficiency of the ICE or
and particulate matter (PM). Manufacturers of Spark Ignition (SI) moving to more costly hybrid drivetrains [3]. Of interest, many
researchers recognize that Waste Heat Recovery (WHR) from
engine exhaust has the potential to decrease fuel consumption
* Corresponding author. Tel.: þ785 864 3181; fax: þ785 864 5254. without increasing emissions, and recent technological advance-
E-mail address: csprouse@ku.edu (C. Sprouse). ments have made these systems viable and cost effective [4].

1359-4311/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.applthermaleng.2012.10.017
712 C. Sprouse III, C. Depcik / Applied Thermal Engineering 51 (2013) 711e722

The two primary sources of waste heat from an ICE are the
engine exhaust (medium-grade) and engine coolant (low-grade).
Other options for heat recovery include the relatively smaller
amounts available from an Exhaust Gas Recirculation (EGR) cooler
and Charge Air Cooler (CAC) [2]. Despite both primary sources
having similar energy content, the higher temperature of the
engine’s exhaust makes it more thermodynamically attractive
when viewed from the perspective of exergy. This results in
a higher theoretical efficiency gain when coupled to a heat engine
[1]. Previous research in this field involves the recovery of each
source, as well as the potential for simultaneous heat recovery from
both the engine coolant and exhaust. Most studies choose
a Rankine cycle for WHR due to its simplicity and ability to operate
with low to moderate temperature differences [5].

1.1. Rankine cycle

A Rankine cycle is a closed-loop system where a working fluid

repeatedly circulates through four components to transform
waste heat into mechanical or electrical power. If the selected
working fluid is organic in nature, researchers often refer to this
system as an Organic Rankine Cycle (ORC). The first step in this
cycle uses engine exhaust to heat the fluid in an evaporator in
order to capture waste heat available (process 2e3 in Fig. 1a).
Selection of the evaporation process as the first step has been
made in order to illustrate the conversion of waste heat to usable
energy; however, the choice is otherwise arbitrary since the
system is cyclic (thermodynamic textbooks primarily indicate
the beginning of this process with the “2” nomenclature). The
superheated vapor then passes through an expander to generate
mechanical or electrical power; this device can be either
a displacement or a turbomachine (process 3e4 in Fig. 1a). After
the fluid undergoes expansion, it flows through a condenser in
order to discharge low-grade thermal energy to the atmosphere
(process 4e1 in Fig. 1a). Finally, a pump compresses the working
fluid back to its original state (process 1e2 in Fig. 1a). To note,
there are two variations of the traditional Rankine cycle; reheat
and regenerative.
In the reheat version, the working fluid does not expand fully to
the condenser pressure in a single stage (process 3e4 in Fig. 1b).
Instead, the cycle sends the partially expanded working fluid back
to the evaporator for reheating (process 4e5 in Fig. 1b) with
a subsequent a second expansion resulting in the working fluid
ending at the condenser pressure (process 5e6 in Fig. 1b). The
principle advantage of the reheat Rankine cycle is an increased
quality at the expander exit [6]. A higher working fluid quality
indicates a reduction in moisture content, which increases the
lifetime of turbine expanders. In the regenerative variation,
a portion of partially expanded working fluid (process 5e6 in
Fig. 1c) preheats the condensed liquid before it enters the boiler
(process 2e3 in Fig. 1c). Preheating in this manner decreases the
amount of heat added at low temperatures, increasing the mean Fig. 1. Variations of Rankine cycle: (a) Traditional, (b) reheat, and (c) regenerative [7].
effective temperature of heat addition while enhancing cycle effi-
ciency [7]. Of pertinence, these explanations only apply to tradi-
tional steam Rankine cycles, not to ORC’s; however, it is important a regenerator and a cylinder, which contains both a displacement
to understand the fundamentals of these cycles before expansion to and power piston [8]. More recently, increased focus has been
ICE utilization. Fig. 1 provides a comparison of the three Rankine placed on the development of Kalina cycle systems, which parallels
cycle variants for clarity of description. an ORC in configuration with the addition of an absorber and flash
tank [9]. This cycle uses a variable composition mixture of
1.2. Alternative WHR methods ammonia and water as the working fluid. Similarly, supercritical
carbon-dioxide systems have drawn attention in various WHR
It is also possible to achieve WHR using other unique thermo- applications. While most WHR research focuses on thermody-
dynamic cycles. One such method is the open Brayton cycle, which namic cycles, thermoelectric (TE) devices offer a unique alterna-
requires only three components. Another common option is the tive, since they directly convert thermal energy into electrical
Stirling cycle engine that includes a closed system comprised of energy.
 C. Sprouse III, C. Depcik / Applied Thermal Engineering 51 (2013) 711e722 713

1.3. WHR comparison from the power cycle creates an inherent emissions advantage.
Since an engine’s emissions are contingent on the combustor
As previously elucidated, an ORC is the focus of most small-scale geometry and its operation, optimization of the combustor design
WHR efforts due to its simplicity and ability to operate efficiently separately can minimize emissions, regardless of power extraction
between small to moderate temperature differences. Another concerns. In order to determine whether sufficient power output
primary advantage of the ORC is the use of widely available and was possible required the creation of computer models and
affordable components because of the similarities between ORC a regenerative Rankine engine prototype.
and refrigeration systems. While no single WHR method is superior The group used a 145.5 hp expander output for component sizing
for every system size and waste heat source, the authors believe and determined the system would fit within a 1972 Ford Galaxie
ORC’s provide an attractive combination of efficiency and afford- engine compartment with only minor modifications. They selected
ability for engine exhaust WHR. Some selected research efforts on Fluorinol-85, 85% tetrafluoroethanol and 15% water by volume, as
alternative WHR methods are described in the following para- the working fluid because of a low freezing point ( 82  F), good
graphs simply to provide context and starting references for the thermal stability, nearly isentropic saturated vapor line, and
reader, since an exhaustive comparison is outside the scope of this acceptable cycle efficiencies for boiler temperatures of 500e625  F.
paper. Additional resources on Kalina cycles [10e13], thermoelec- The prototype used a radial firing burner integrated with the vapor
trics [14e16] and carbon dioxide transcritical cycles [17e19] are generator to evaporate the working fluid that generated power
available. within the expander. Tests using a single and four-cylinder (V-4)
By contrast, relevant studies have found that various Brayton expander found that the single cylinder expander exhibited supe-
cycles are marginally attractive as an alternative to turbo- rior efficiency. They observed good agreement between computer
compounding [20]; however, they require a longer payback time models and prototypes and concluded that an 11% fuel economy
[21], and are less attractive for low to medium temperature WHR improvement was achievable based on 400 h of testing utilizing
[22]. Similarly, Kubo found that several Stirling cycle configurations a 5.5 hp system at Thermo Electron Corporation (TECO).
to recover waste heat from an adiabatic diesel engine were not Three years later, Patel and Doyle documented the first appli-
economically attractive [8], and fail to match previous perfor- cation of using an ORC for engine WHR [36]. Their system used the
mances of ORC [23] and steam Rankine cycle [24] systems. More exhaust waste heat of a Mack 676 diesel engine installed in a long
recently, the ORC investigated by Bianchi and De Pascale was haul truck by coupling the turbine pinion gear to the PTO with
favored over the Stirling engine [22]. Kalina cycle research has a speed reduction gearbox. While small vehicles with ICE’s have
focused primarily on large-scale facilities, demonstrating that these potential for efficiency gains, the authors cite extended periods of
systems efficiently recover waste heat in the engine exhaust near constant engine speed operation and high yearly mileage as
temperature range [9,25e28]. Practical restrictions on maximum reasons long haul trucks are more attractive for ORC applications.
cycle pressure suggest that carbon dioxide transcritical cycles To minimize irreversibility, the group used Fluorinol-50 as the cycle
would not be feasible for exhaust WHR, despite the cycle’s favor- working fluid in an attempt to minimize the temperature difference
able performance shown by Chen at low temperatures [29]. between the working fluid and engine exhaust. In specific, they
A summary of the present and future applications of thermo- operated between 650  F at turbine inlet and 158  F at condenser
electric devices is provided by Riffat and Ma, defining their exit.
maximum temperature difference as 70  C and efficiency at around The constructed prototype used a vapor generator as part of the
5% [30]. Even with cascading devices, which will function for exhaust stack, a condenser combined with the existing truck’s
differentials of 130  C, the authors restrict the competitive regime radiator, and a three-stage axial flow turbine expander. Figur-
of TE’s to low temperature waste heat. If efficient WHR is not es within their report illustrate that the ORC components were
required to achieve the objective, such as meeting the electrical readily configured within the truck’s engine compartment. Sub-
requirements of a family car, TE’s could provide a viable alternative jecting the ORC equipped truck to a NAPCA control route demon-
to ORC systems [31]. strated a 13% increase in maximum power output along with a 15%
The next section will focus on a historical review of ORCs for improvement in fuel economy. Moreover, this performance slightly
WHR and power generation. Of importance, in addition to ICE exceeded model predictions at all power levels and speeds tested.
waste heat, Najjar recognizes the possibility of using ORC’s for WHR They conclude that the $3000 ORC system could pay for itself with
from other power generating devices, such as gas turbines [32]. fuel savings in less than one year with 150,000 miles of travel at the
More recently, Quoilin and Lemort described the use of ORC’s as current fuel cost of $0.45 per gallon.
part of several green energy systems [33]. Among these were Continuing the collaboration between TECO and Mack Trucks,
biomass, solar, and geothermal stations. Both works include Patel et al. describe plans for a yearlong single vehicle test of the
descriptions of previous publications in their respective research ORC equipped model F7865T tractor in FY 1979 [37]. This paper also
areas. However, the primary goal of the authors is to increase the documents a planned expansion of the system for ten trucks in the
thermal efficiency of the ICE for lower greenhouse gas emissions FY 1981-1982; however, no further information exists on the
without adversely influencing emissions. Of interest, there has planned expansion. Using the previously determined 15% fuel
been research on using recovered exhaust waste heat for automo- consumption decrease, they mention saving 3120 gallons of fuel for
bile heating, offering a secondary avenue of usage [34]. every 100,000 miles traveled by the truck.
In 1981, Hall summarizes the increase in cogeneration, which he
2. WHR review describes as the generation of energy from the excess energy
supplied during another process [38]. This necessity stems from the
In 1973, Morgan et al. were among the first to study the appli- 100% OPEC inspired price hike on oil, which made it less expensive
cation of an external combustion Rankine engine in an automotive to save a Btu than to generate an additional one. While concerns
application [35]. Concerns about the ability of ICE’s to meet the about the substantial capital investment still existed at the time of
Clean Air Act mandated emission levels for 1976 motivated these the paper, Hall contends that installation of cogeneration systems is
researchers to investigate alternatives to traditional combustion not overly complex.
engines. As a result, the group evaluated the potential of an ORC In the same year, Heywood provides a review of the status of
system as they recognized that separating the combustion process conventional engines as compared to alternative versions [39].
714 C. Sprouse III, C. Depcik / Applied Thermal Engineering 51 (2013) 711e722

During this time in the literature, the efficiency improvement three superior candidates. While none completely satisfied all the
potential of direct-injection and turbo-compounded diesels were criteria, the short listed fluids were the fluorinated hydrocarbons R-
being tested for future incorporation. The alternative systems 11, R-113, and R-114.
identified by Heywood were gas turbine systems, Stirling engines, After considering a turbine or displacement-type machine for
and Rankine cycle engines. This paper describes the advantages and the Rankine engine expander, Badr chose to use a multi-vane type
disadvantages of each system; however, he did not intend the displacement machine. At this time, single-stage turbines were
study to be definitive because of rapidly changing constraints. Of inefficient for this application due to impractical optimum blade
interest, Heywood acknowledges using an ORC for WHR as velocity. Moreover, the high cost of multi-stage turbines prohibited
a possible method for further efficiency improvements over turbo- their inclusion. Computer modeling of the preferred fluids revealed
compounded diesels. R-11 exhibited the highest cycle efficiency; however, since R-11 has
Also in 1981, Marciniak compares the suitability of seven a relatively low thermal stability limit of 120  C, this work
different working fluids for Rankine cycle systems [40]. The fluids concludes with R-113 as the best working fluid.
considered were water, methanol, 2-methyl pyridine/water, Also in 1985, Bailey performs a technical and economic study of
Fluorinol-85, toluene, Freon R-11, and Freon R-113. This working three alternative power cycles for WHR from an adiabatic diesel
fluid evaluation occurred for 600e2400 kW industrial applications engine, and compares each system’s performance with a baseline
with waste heat temperatures between 500 and 1100  F. Marciniak turbo-compounding diesel [21]. The power cycles examined were
makes these working fluid recommendations after studying their a steam Rankine, ORC with RC-1 working fluid, and various Brayton
thermo-physical properties, system performance and size, envi- cycles (including air Brayton). Bailey documents engine specific fuel
ronmental impact and safety, various capital and operating costs, economy improvements via computer models, which are in direct
and potential return on investment. He does not exclude any of the proportion to improvements in vehicle fuel economy [46]. Simula-
potential working fluids due to health or safety risks, believing these tions show the largest fuel savings were possible with the RC-1 ORC,
concerns were manageable through a careful system design. Under followed closely by the steam Rankine, with smaller improvements
high temperature conditions, thermal stability concerns eliminated from Brayton cycle systems. The calculations were based on
methanol and both Freon refrigerants from consideration. From an 100,000 miles of annual travel, with the RC-1 ORC providing a fuel
economic standpoint, this paper provides the estimated capital cost savings of approximately 1400 gallons, the largest among the
versus power output for the distinctive working fluids. As power different options. At 1985 fuel price levels, the steam Rankine cycle
output increases from 800 kW to 2400 kW, typical Rankine system has the shortest payback time at 3.2 years, followed by the RC-1 ORC
costs are shown to exponentially decay from around $1000/kW to and intercooled Brayton at around 4.1 and 4.7 years respectively.
$600/kW. The least expensive system utilized water, followed by Six years later, Badr elaborates on his 1985 paper by ranking
toluene, 2-methyl pyridine/water, and Fluorinol-85. types of low power output (5e20 kW) expansion machines for use
In the next year, Poulin et al. conduct a similar evaluation in in a steam Rankine engine configuration [5]. A Rankine cycle fits
running a steam Rankine cycle system off several adiabatic diesel this application because it is able to achieve 60e70% of the Carnot
configurations [24]. They simulate this system using turbocharged limit for small temperature differences; whereas, it can only ach-
non-aftercooled (TC), turbocharged aftercooled (TC/A), turbo- ieve up to 50e60% for large differentials in temperatures. He
charged turbo-compound non-aftercooled (TCPD), and turbo- evaluates these devices for a combined heat-and-power (CHP)
charged turbo-compound aftercooled (TCPD/A) engines. They find system, which involves the simultaneous production of electrical
a maximum brake specific fuel consumption (BSFC) improvement power and useful heat from a single generating power plant. A
of 16.2% for the TC diesel with lesser improvements for the other survey of previous research led the author to conclude that turbines
configurations. They conclude that the payback time including and reciprocating expanders were not particularly suitable for
maintenance would be 2.3 years, making the system economically a low power, low speed application. Badr concludes that a Wankel
attractive to the end user. expander was best for the CHP system, slightly above a helical-
Also in 1984, Italian research in the field of ORC’s was described screw expander.
by Angelino et al. [41]. This review of the authors’ works includes 14 Jumping ahead to 1993, Oomori and Ogino develop a Rankine
ORC engines with power outputs between 3 and 500 kW, and cycle system that uses the waste heat from the engine coolant in
maximum operating temperatures between 70 and 340  C. Of the a passenger car [47]. The authors chose not to recover the higher-
studies featuring higher maximum cycle temperatures, high grade waste heat found in the engine’s exhaust because engine
molecular weight wet (negative saturated vapor curve slope) fluids speed fluctuations complicate the control process. They select
are used to generate power through multi-stage turbines. While HCFC123 (hydrochlorofluorocarbon) as the working fluid for this
relatively few cycle efficiency results are presented, the authors application in order to allow liquidevapor phase change within the
show a small increase in performance as the working fluid’s authors’ low temperature evaporator. Moreover, they replace the
molecular weight increases. engine water pump with an ORC pump, while using the engine
By 1985, increased interest in simulating ORC’s led Badr to coolant passages as the evaporator, and the existing radiator as the
develop correlations for thermophysical properties of many of the condenser. They measure output from the scroll expander in order
period’s refrigerants [42]. Badr bases these correlations off works to determine the fuel economy improvement during bench testing
by Martin [43] and Downing [44], while showing good agreement and find only a 3% gain for an ambient temperature of 25  C.
with the limited data available. This effort provides equations that Two years later, Larjola evaluates prototype high-speed oil free
can fit directly in a model for a working fluid’s saturation pressure, turbogenerator-feed pumps in an ORC for WHR from a 425  C
specific volume, specific enthalpy, and specific entropy. Of impor- source [48]. Instead of using the conventional separated configu-
tance, the correlations hold for each refrigerant as a saturated ration, the novel device directly couples the pump, turbine, and
liquid, saturated vapor, and superheated vapor. generator. Upon examination of several working fluids, toluene is
Badr publishes another paper in the same year describing the chosen as the most suitable. The author notes the fluid’s lower
working fluid selection process for a low power output (<10 kW) specific heat of vaporization and the possibility of using a single-
Rankine cycle engine operating between 40  C and 120  C [45]. This stage turbine as advantages toluene has over water. This is
paper includes an evaluation of thermodynamic and physical further indicated in the results as a toluene-based ORC achieves
selection criteria for 67 prospective working fluids, while identifying 26% efficiency while a steam Rankine cycle manages only 11e19%.
 C. Sprouse III, C. Depcik / Applied Thermal Engineering 51 (2013) 711e722 715

In 1997, Hung compares the performance of six working fluids In the next year, Liu et al. conduct an analysis of working fluid
against water for use in a Rankine cycle for WHR of low-grade selection on the performance of an ORC for WHR [52]. They model
waste heat [49]. The candidate fluids were benzene, ammonia, a simple ideal Rankine cycle in order to study the thermal efficiency
R-11, R-12, R-134a, and R-113. Hung chooses water as the baseline of the system and total heat-recovery efficiency for various working
fluid since it provides low Rankine cycle efficiency for waste heat fluids utilizing low-grade energy sources. For completeness, the
sources below 370  C; thus, a direct comparison of organic fluids to authors choose to evaluate wet, dry, and isentropic fluids in the
water justifies the use of organic fluids in low temperature appli- study. The authors determine that hydrogen bonding in water,
cations. Determination of working fluid suitability is a function of ammonia, and ethanol results in wet fluids that are inappropriate
the examination of molecular weight, saturated vapor line slope, for a low temperature ORC process. Simulation results predict that
turbine enthalpy drop, maximum stability temperature, and critical thermal efficiency is a weak function of the critical temperature of
point as indicated in Table 1. This paper involves a modeling study the fluid and lower critical temperature fluids provide less efficient
in order to determine the maximum Rankine cycle efficiency for heat recovery.
each fluid at different turbine inlet temperatures. As shown in In the same year, Schmid examines potential CO2 emissions
Fig. 2, benzene has the highest efficiency of the investigated fluids reductions by utilizing WHR on marine engines [53]. Since these
in spite of having a relatively high latent heat at low pressure. In engines achieve around 50% thermal efficiency along with the
general, the larger latent heat working fluids result in inefficient typical tradeoff between efficiency and NOx emissions, the author
operation since the condenser rejects much of the fluid energy. In cites WHR as a more promising avenue for CO2 emission reductions
addition, for given evaporator and condenser temperatures, higher than engine redesign. Schmid proposes a heat recovery plant that
critical temperature working fluids exhibit superior first-law effi- generates electricity using a steam turbine and power turbine in
ciencies in an ORC. order to drive a turbo-generator. Their calculations indicate that the
Hung follows up this earlier work four years later by examining operation of the heat recovery plant by means of a 68,640 kW
the potential of five different dry fluids (fluids with a positive slope Sulzer 12RTA96C engine results in a 12% gain in overall efficiency.
on their Tes diagram saturated vapor curve) to perform WHR in an Because the estimated payback time was only four years, the
ORC [50]. This investigation includes three hydrocarbons and two manufacturer went ahead and installed the complete heat recovery
refrigerants as follows: benzene (C6H6), toluene (C7H8), p-xylene plant on some ships in the manufacturer’s fleet [54].
(C8H10), R113 and R123. Using a 10 MW waste heat source at 600 K, In 2005, El Chammas and Clodic propose an ORC for WHR from
Hung finds that p-xylene exhibits the lowest irreversibility of the the cooling circuit and exhaust of a 1.4-L spark ignition engine
prospective fluids. Another study, using a constant 15  C temper- within a hybrid vehicle [4]. With this configuration, conversion of
ature difference between the turbine inlet and waste heat source, mechanical turbine rotation to electrical energy with a generator
shows that p-xylene provides the highest cycle efficiency. The allows subsequent storage in a battery pack for later use by an
refrigerants perform relatively better with decreasing source electric motor. The authors examine eight working fluids (water,
temperature; however, using these relatively expensive fluids in isopentane, R-123, R-245ca, R-245fa, butane, isobutane, R-152a) for
a large system would add significant cost. use in a Rankine bottoming cycle. The simulation results in Fig. 3
In 2003, Kane et al. propose creating a small hybrid power utilize a condensing temperature of 55  C while indicating that
generator utilizing WHR for remote locations [51]. This station water provides the highest cycle efficiency, followed by R-123,
would recover thermal energy from solar collectors and a diesel isopentane, and R-245ca. In spite of R-123 resulting in efficient
engine while creating electricity through two superimposed ORC’s. operation, the authors note the fluid is an HCFC and thus unsus-
This design allows power generation during cloudy periods and at tainable for environmental reasons.
night via WHR from the engine block coolant and exhaust gas. The Their efforts find that dry or isentropic fluids achieve similar or
ORC’s involve a topping cycle with HCFC-123 and a bottoming cycle lower efficiencies with increasing superheat at the turbine inlet,
using HFC-134a with power generation through a separate scroll while the cycle efficiency increases for wet fluids such as water.
expander unit. While a single definition of topping and bottoming Rankine cycle efficiency does increase for higher operating pres-
will not apply to all the papers reviewed here, in general, a topping sures; however, they mention that manufacturers must weight
cycle will convert waste heat to mechanical or electrical energy; these benefits against the associated cost increase. A comparison of
whereas, a bottoming cycle redirects the waste heat to a useful possible Rankine cycles, including a high-pressure steam Rankine
heating process. Earlier in this paper, Khalifa indicates the nomen- cycle (HPSRC) and low-pressure steam Rankine cycle (LPSRC) show
clature of a bottoming cycle in reference to a Brayton cycle gener- that the HPSRC had the highest actual efficiency as indicated in
ating power from engine exhaust waste heat. For Kane et al.’s paper, Table 2. In particular, HPSRC offers a 12e27% fuel economy
the authors associate the topping cycle with the higher exergy improvement depending on engine speed and load, while the two
source (exhaust heat) and the bottoming cycle with the lower exergy favored organic fluids (isopentane, R-245ca) present a 17e32%
sources (engine coolant and heat rejected from topping cycle improvement when used in an ORC. Additional exhaust
condenser). To test efficiency, they ran three series of tests using the backpressure introduced by the boiler dampens this performance
station including limited testing made in situ with the fully inte- increase, resulting in a horsepower loss of 1.5e2.5% at high loads
grated system. Results show an overall cycle efficiency of 14.1  0.2% and 0.2e0.5% at low loads.
with the performance deemed satisfactory for relatively low In 2006, Stobart and Weerasinghe review and compare six
temperatures (up to 165  C) and low power ranges (up to 10 kW). previous WHR efforts for SI and CI engines with a summary of the

Table 1
Thermophysical working fluid properties [49].

Parameter H2O NH3 Benzene R134a R12 R11 R113

Molecular weight 18 17 78.14 102 121 137 187
Saturated vapor curve slope Negative Negative Positive Isentropic Isentropic Isentropic Positive
Critical point 647 K 22.06 MPa 405.3 K 11.33 MPa 562.2 K 4.9 MPa 374.15 K 4.06 MPa 385 K 4.13 MPa 471 K 4.41 MPa 487.3 K 3.41 MPa
Latent heat at 1 atm (kJ/kg) 2256.6 1347 438.648 215.52 166.1 178.8 143.9
Max. stability temperature (K) None 750 600 450 450 420e450 450e500
716 C. Sprouse III, C. Depcik / Applied Thermal Engineering 51 (2013) 711e722

Table 2
Comparison of Rankine cycles [4].

Parameter HPSRC LPSRC ORC-R245ca ORC-isopentane

Net electrical output (kW) 5 5 5 5
Ideal efficiency (%) 22.7 15.3 15.8 16.3
Actual efficiency (%) 13.3 8.9 8.7 9.0
Max pressure (bar) 25 8 22.5 19
Max temp. ( C) 260 220 143 151
Mass flow rate (g/s) 15.2 22.7 241 121

Also in 2006, Leibowitz et al. highlight the advantages of using

screw expanders over turbines in small (20e50 kW) ORC’s by
considering both cost and performance [57]. At these sizes, the
authors note the high rotational speeds of turbines require high
ratio gearboxes and expensive lubrication systems. Of the different
types of expanders, twin-screw type machines are identified as the
Fig. 2. Variation in system efficiency for various working fluids [29].
most promising, citing their ability to operate at higher rotational
speeds than vane, scroll, or reciprocating expanders. Consisting of
a pair of meshing helical rotors, recent development of the “N”
considered studies [55]. Four of the six works recover heat from
profile has improved the screw expander’s performance by up to
exhaust gas, with operating efficiencies between 9 and 25%. They
10%. Furthermore, the twin-screw expander provides efficiencies
construct two QSS-Toolbox computer simulation models for anal-
around 70% during low speed operation (1500e1800 RPM),
ysis of a steam hybrid system based off a 1.6 L VW Golf car while
resulting in acceptable performance during periods of decreased
simulating the performance of the WHR system over several drive
waste heat supply. Cycle simulations showed an 8.5% increase in
cycles. They find fuel consumption decreases of 9.6%, 8.9%, and
total power output by recovering exhaust heat from a gasoline
26.2% respectively for the European, FTP 75, and US06 drive cycles.
engine generator at a cost of $1500e1800/kW.
The lower fuel consumption of the US06 drive cycle can be attrib-
Colonna and van Putten publish a two-part effort in 2007,
uted to the greater loads placed upon the engine during higher
describing a dynamic power cycle simulation software package
speed (highway) driving resulting in more exhaust energy poten-
called SimECS [58,59]. This software uses conservation (mass,
tial. In addition to the simulation results, the authors identify high
momentum, energy) and constitutive (heat transfer coefficients,
efficiency as the main attribute of turbine expanders, while recip-
fluid properties, fluid dynamic correlations, rate of reaction)
rocating expanders offer superior response to fluctuating exhaust
equations in order to describe the complete system. The use of
conditions.
bilateral coupling and the causality principle leads to a solvable
In the same year, Arias et al. consider three Rankine cycles for
system of low index algebraic and differential equations (DAEs).
WHR from a spark-ignition engine in a hybrid vehicle [56]. The
Validation of the software occurs through steady state and dynamic
configurations include heat recovery from exhaust gas only, engine
simulation of a 0.6 MW steam Rankine system in the second part of
coolant þ exhaust gas, and engine block þ exhaust gas. The authors
the work. Results indicate a maximum discrepancy of 6% between
use an energy balance to quantify the amount of waste heat
Cycle-Tempo and SimECS computer simulation models. These
available in each configuration and couple this with experimental
works present a highly accurate method of modeling WHR systems
data from a 2004 Toyota Prius in order to form the basis of their
during normal and off-design condition operation, representing
computer simulations. These models show that using the engine
one of the first detailed treatments of transient cycle simulation.
block to preheat the working fluid before superheating the steam
In the same year, Wei et al. investigate the performance of an
from exhaust energy was the most efficient configuration, con-
ORC system during steady operation and under disturbances [60].
verting 7.5% of the total waste heat to electrical power.
Exhaust from a gas turbine (610e650 K) acts as the waste heat
source driving the cycle, which uses R-245fa as the working fluid
and a separate turbine to generate power. This work validates cycle
simulations on a 100 kW system through obtained experimental
data, ensuring accurate model predictions. Optimization of the ORC
performance suggested small sub-cooling (0.5e0.6 K) at the
condenser outlet leads to the maximum WHR system output.
In this study, the authors quantify the effect on cycle perfor-
mance of variations in exhaust mass flow rate, exhaust inlet
temperature, air mass flow rate, and ambient temperatures. These
simulations show linear increases in output power and efficiency
for increasing exhaust mass flow rate, increasing exhaust inlet
temperature, and decreasing ambient temperature. They observe
a 30% decrease in output power from typical winter temperatures
as compared to those of summer because of the higher cold sink
temperature. Finally, the exergy destruction rate of the evaporator
is the largest of all the ORC components in the cycle.
Also published in 2007 is a two-part effort by Teng et al.
regarding WHR from heavy-duty diesel engines using an ORC
[2,61]. The authors propose a hybrid energy system for 2010
Fig. 3. Rankine cycle efficiency for various working fluids [4]. versions of these engines in the first paper, consisting of a diesel
 C. Sprouse III, C. Depcik / Applied Thermal Engineering 51 (2013) 711e722 717

cycle hybridized with an ORC for WHR. Transient operating This year additionally provides the efforts of Endo et al. with
conditions commonly lead to selecting reciprocating expanders respect to their experimental results involving a steam Rankine
above turbines, since variable heat addition leads to wetness late in system installed on a Honda Stream hybrid vehicle [1]. The WHR
the expansion process and potential damage to turbine blades. The power cycle utilizes the waste heat from engine cooling and engine
working fluid selected for the cycle is a dry fluid having a critical exhaust using an innovative catalyzed evaporation device that
pressure under 70 bars. Both first and second law analyses reveal replaces the exhaust port of the engine. The system generates
that WHR from engine coolant was unattractive due its respective power from a swash plate axial piston type expander coupled to
low energy level; however, their investigation indicates that energy a generator. Electronic control of the generator load provides steam
from engine exhaust, EGR, and the CAC is suitable. The authors pressure regulation by imposing the rotational speed of the
provide a schematic of the proposed system, where they place an expander. The authors use feed-forward control of the steam flow
exhaust cooler downstream of a Diesel Particulate Filter (DPF). rate in order to maintain the expander inlet temperature between
Their simulations indicate that the supercritical ORC system can 400 and 500  C.
potentially recover 55 kW from the waste heat (169 kW) via the For proof of concept, they drive the vehicle in the EPA highway
three devices, increasing the output of the engine by 20% without fuel economy test mode, Japanese 10e15 mode, and at constant
additional fuel. A further benefit in their proposed cycle is achiev- driving speed of 100 km/h. They observe successful control of the
able by decreasing the condenser size through utilizing heat from system during broad and dramatic engine load changes, the first
engine coolant. reported instance of such transient control capabilities. Experi-
Their second paper discusses the merits of various working mental data shows the novel evaporator recovers 95% of the
fluids for their ORC-WHR system based on thermodynamic prop- exhaust energy as steam and heat from the catalytic purification
erties. This paper documents inorganic and organic fluids as both reaction, while the expander achieves a maximum of 13% thermal
pure fluids and binary-mixture fluids. In specific, characterizations efficiency. In addition, results show a 13.2% increase in thermal
of these fluids happen as wet (water), isentropic (R134a), or dry efficiency relative to the base vehicle, leading the authors to
(R245fa). Of these, the authors determine isentropic, slightly wet, conclude that the Rankine system is an effective means for
or slightly dry fluids are generally most suitable for ORC-WHR improving vehicle fuel economy.
systems. However, selection of the proper working fluid depends Stobart et al. continue advancing the control literature with
on the temperature range of the system; hence, no single fluid will their effort on regulating thermal recovery systems in 2007 [63].
be best for all applications. The authors also conclude a properly The lower manufacturing costs and high exhaust temperatures lead
selected binary-mixture working fluid is capable of improving the the authors to prefer spark ignition engines to compression ignition
thermal efficiency of the ORC by up to 15%. engines for vehicle implementation. Separation of thermal
In the same year, Mago et al. investigate the performance of low recovery from work production with a pressurized accumulator
temperature ORC’s with seven working fluids [62]. The fluids and hybrid architecture combat the unsteady waste heat condi-
considered were R134a, R113, R245ca, R245fa, R123, isobutane, and tions, improving controllability. Heating times for the 50 kW engine
propane. They compare the performance of each fluid against water example, achieving heat recovery with a steam Rankine cycle
in waste heat temperature ranges such as less than 380 K, 380 to featuring an accumulator, ranges from about three to four min.
430 K, and greater than 430 K. As mentioned in a previous para- Reduced heating times are achieved with smaller water volumes
graph, water is not economical for waste heat levels below and after short stops. Of the expander options, reciprocating
a temperature of 643 K. Fig. 4 presents the results of their efforts as expanders are relatively efficient and simple to control.
a function of computer simulated ORC thermal efficiency with each Concurrently in 2007, Lemort et al. presents an experimentally
fluid. This allows the researchers to make recommendations for validated numerical model of an ORC using HCFC-123 as the working
each temperature range. In specific, the working fluid with the fluid [64]. This effort links individual sub-models of each Rankine
highest boiling point (R113) provides the highest efficiency, while cycle component in order to create an overall model, allowing
they observe that the worst efficiency results from the fluid with optimization of the cycle operating conditions for maximum
the lowest boiling point (propane). This suggests that higher performance. They utilize a scroll expander sub-model from
boiling point fluids work comparatively better in ORC’s. a previously published work [65] that uses steam as the working
fluid. As a result, the model requires recalibration of certain
parameters for HCFC-123. Heat enters the cycle from two hot air
streams (177e189  C and 137e160  C) in an amount around 15 kW,
while converting between 0.7 and 1.0 kW into mechanical power.
Experimental data from 13 tests is presented for comparison with
simulations, validating the accuracy of the global cycle model. The
authors’ model is shown to agree relatively well with experimental
measurements in spite of involving a limited number of parameters.
In what continues as a busy year, Quoilin describes experimental
results and optimization of a small scale ORC through simulations
of a validated computer model [66]. After considering numerous
prospective working fluids, selection of an isentropic fluid (R-123)
happens because of its efficient operation with source tempera-
tures between 100 and 200  C. The authors present a schematic of
their ORC test bench, which utilizes a scroll expander modified
from a scroll compressor. Achieving maximum WHR potential of
the system occurs by varying certain parameters such as the hot air
source temperature, expander rotational speed, and refrigerant
charge. Results demonstrate maximum cycle efficiency of 7.4%,
Fig. 4. Thermal efficiency of Rankine cycle with turbine inlet temperature for various while simulations accounting for recommended modifications
working fluids [46]. predict a further efficiency gain up to 11.58%.
718 C. Sprouse III, C. Depcik / Applied Thermal Engineering 51 (2013) 711e722

This same year, Kadota and Yamamoto develop a transient cycle, while selecting a rotary-vane type expander in order to
bench for simulation of Honda vehicles with various power-train minimize system costs. Laboratory test results reveal a maximum
configurations [3]. Among the simulated power plants were thermal efficiency of 3.82% for an 80  C temperature difference and
conventional IC engines, hybrid electric (HE), and HE with a steam an expander efficiency of 40%.
ORC for WHR. Schematics in the paper present the components Additionally in this year, Espinosa et al. discuss the optimal ORC
utilized in each model, which the researchers validated experi- configuration for WHR on commercial trucks [71]. This work
mentally in an individual manner before evaluating system considers heat sources from the CAC, engine coolant, EGR, and
performance over a full driving cycle. The HE system with WHR exhaust gases. They mention that exhaust gases are the most
demonstrates a 6.1% thermal efficiency improvement in the Japa- practical waste heat source since simultaneous WHR from both
nese 10e15 mode. Fuel economy predictions were within 1 EGR and the exhaust line adds complexity and increases cost.
percent of measured values. Moreover, the energy available in engine coolant is of low quality.
In 2009, Ringler et al. investigate the use of an ORC on a four- The authors prefer using the ram air speed from vehicle motion in
cylinder engine for WHR [67]. They examine system performance order to cool the ORC condenser since the power consumption of
as a function of heat recovery from exhaust only (system A) and a fan would increase with the square of the required airflow rate.
exhaust þ engine coolant (system B). Results from this effort This work uses computer models to evaluate water, ethanol, and
determine that water was the most appropriate fluid for system A HFC-R245fa with the latter chosen as the most appropriate working
while an alcohol (e.g. ethanol) was best for system B due to a lower fluid. The high global warming potential of HFC-245fa suggests an
temperature heat addition. To predict additional performance, HFO fluid with higher critical temperature than HFO-1234yf could
simulations using a Dymola modeling tool examine the different be promising in future Rankine cycles. Finally, they simulate the
systems with A having better performance when vehicle speeds different ORC configurations using zero-dimensional and one-
exceed approximately 65 mph. Alternatively, system B demon- dimensional simulation models. The zero-dimensional model
strates a higher potential for typical driving between 45 and 70 uses isentropic efficiencies and three-zone heat exchangers;
miles per hour, resulting in an engine performance increase in the whereas, the one-dimensional model comes from the commercial
range of 10%. engine simulation program GT-POWER.
Use of an ORC to recover waste heat from a high pressure ratio Meanwhile, three possible ORC configurations are considered by
gas turbine studied by Chacartegui et al. is also accomplished in Vaja and Gambaratta in 2010 in order to recover waste heat from
2009 [68]. While the heat recovery is not achieved using an ICE as a 12 cylinder supercharged natural gas engine [72]. These candi-
the power plant, the gas turbine exhaust temperature is compa- dates are a simple ORC using exhaust gas, a simple cycle using
rable, making the results applicable to the current effort. Further- exhaust and coolant, and a regenerated cycle. Each of the three
more, this work is included because it compares the six potential typical classes of working fluids are represented by benzene (dry),
working fluids of R113, R245, isobutene, toluene, cyclohexane, and R11 (isentropic), and R134 (wet). Environmental and safety
isopentane. Parametric optimization of the ORC, along with several concerns are recognized in the paper; however, the authors
commercial gas turbines, is used to predict the overall system conclude their results only based on thermodynamic attributes.
performance. Of the fluids, cyclohexane and toluene provide the The regenerative cycle, as well as the simple cycle using both
highest global efficiencies, with comparable system costs. exhaust and coolant waste heat along with benzene as the working
That same year, a comparative study of four potential methods fluid, provided the highest efficiency improvements over the
for waste heat recovery in cement plants is done by Wang et al. engine rated value at around 12.5%. However, the simple cycle
[28]. Two sources are used to power the thermodynamic cycles, the using just exhaust energy and benzene has a comparable efficiency
preheater (340  C) and clinker cooler (320  C), both resembling the of 11.4%. As a result, the designer must weigh a slight drop in effi-
low end of ICE exhaust temperatures. The systems evaluated are ciency against a reduction in initial cost and complexity. R11 was
single flash steam cycle, dual pressure steam cycle, ORC, and Kalina noted as the next most attractive fluid, with efficiency increases
cycle. An exergy analysis is examined with genetic algorithm around 10% for both simple cycles.
parameter optimization for each candidate, resulting in the A further effort by Yamada et al. in 2010 describes a unique
conclusion that the Kalina cycle would achieve the best perfor- Rankine cycle configuration for WHR from a hydrogen ICE [73]. The
mance. For comparison, the ORC achieved 20.9% efficiency while authors propose an open Rankine cycle that discharges the working
the more complex Kalina cycle provided 24.1%. fluid to the atmosphere after expansion. In this system, a water
In 2010, Srinivasan et al. examine the waste heat recovery separator recovers water from the exhaust pipe that passes through
potential of an ORC for improving the efficiency of a high-efficiency, a filter and subsequently through the pump, indicating the start of
low-emissions dual fuel Low Temperature Combustion (LTC) engine the cycle. A modified exhaust manifold acts as the cycle evaporator,
[69]. The authors’ single-cylinder test engine uses Advanced Low followed by an expander. The cycle was simulated with and without
Pilot Ignited Natural Gas (ALPING) LTC, where pilot diesel sprays a condenser, leading to the conclusion that the condenser was not
mix with surrounding natural gas prior to combustion. Simulations cost-effective. Inclusion of the open Rankine cycle without
using exergy and pinch-point analysis, along with measured engine a condenser provides thermal efficiency improvements over the
exhaust parameters, predict Fuel Conversion Efficiency (FCE) and hydrogen ICE of 2.9e3.7% at various engine speeds.
emissions improvements. Different injection timing, engine load, A clear example of the progress toward efficient and economical
and EGR percentages are used to find results for a range of engine small ORC’s is presented by Quoilin et al. in the same year [74]. This
operation. On average, an R113 working-fluid based ORC improved work describes a semi-empirical model of an ORC using the
FCE by 7% while reducing NOx and CO2 emissions by 18%. Since R113 refrigerant HCFC-123 along with a previously validated scroll
is a dry fluid, the authors achieve the optimum efficiency by expander model. By modifying a mass produced, oil-free open-
evaporating to the saturated vapor curve, avoiding superheat. drive scroll compressor to operate as an expander, the authors limit
Concurrently, Mohd et al. describe their low power output system cost while achieving heat recovery efficiency equal to 9.9%.
(<1 kW) ORC for energy generation from low temperature waste This was accomplished after optimizing cycle parameters while
heat [70]. This work experimentally evaluates heat sources ranging generating power from two low temperature sources.
from 60 to 100  C and cold source temperatures between 10 and In what continues to mark a busy year, Weerasinghe et al.
30  C. The authors choose HFC-245a as the working fluid for the compare Rankine cycle WHR with turbo-compounding as methods
 C. Sprouse III, C. Depcik / Applied Thermal Engineering 51 (2013) 711e722 719

to boost the thermal efficiency of diesel engines [75]. In effect, temperature and evaporation enthalpy provide the maximum
turbo-compounding directly expands exhaust gases in a turbine for power but requires high boiler pressures, low condensing pressures,
power generation, representing part of a Brayton cycle. An inter- and a high degree of superheating. They find that implementing
esting aspect of the Rankine cycle is the steam storage reservoir a Rankine cycle on heavy-duty applications can increase the total
between the evaporator and expander, which provides an energy powertrain efficiency by up to 30% based on the NEDC driving cycle.
buffer through periods of high-load or low-load engine operation. They state that while developments in EHR have rapidly progressed
While water is chosen as the cycle working fluid, the authors note over the past decade, attempts to apply EHR to production vehicles
that further advancements in steam expanders are underway at the must continue moving forward in order to ensure meeting future
University of Sussex. A Matlab based model using the QSS Toolbox fuel economy and emission standards. This requires demonstration
simulates both systems over the US Federal Heavy Duty Transient of efficient thermal EHR and precise control with small hardware.
Test Cycle, resulting in a fuel economy improvement of 20% or more Just recently, Bianchi and De Pascale evaluated three thermo-
with the Rankine cycle and only around a 2.5% savings from turbo- dynamic cycles in order to recover low and medium grade waste
compounding. heat [22]. These cycles are the ORC, Stirling, and the inverted
One year later, the performance of an ORC using R-12, R-123, Brayton; of which they resolve that an ORC is the most attractive of
R-134a, and R-717 is parametrically optimized by Roy et al. [76]. the candidates. Four organic fluids (R133a, R245fa, iso-butane, and
Both a constant heat source temperature of 550 K and a variable benzene) were considered for the ORC, with benzene achieving the
heat source are considered in the study. In each case, R-123 highest efficiency at around 20%.
produces the maximum efficiency (w19%) and lowest irrevers-
ibility of the candidate fluids. Along with the thermodynamic 2.1. Review summary
analysis of each fluid using Matlab, environmental concerns are
presented with respect to the Ozone Depletion Potential (ODP) and Growing concerns over air pollution and escalating fuel prices
Global Warming Potential (GWP) of the refrigerants. They indicate through the 1970’s resulted in attempts to find viable alternatives
that ammonia (R-717) has zero impact on ODP and GWP, while R-12 to the IC engine. However, the low power density, high cost, and
(a chlorofluorocarbon: CFC) contributes to global warming and low efficiency of other power cycles, prevented a substantial
continually destroys ozone for up to two years. Furthermore, the change in vehicle propulsion. Instead, the focus narrowed to WHR
HCFC studied (R-123) has both a low ODP and GWP; whereas, R- systems in order to increase efficiency through cogeneration by
134a (a hydrofluorocarbon: HFC) does not have any ODP but has supplementing the engine’s power by means of converting thermal
a high GWP. energy into a more useful form; either mechanical or electrical.
While most authors are designing for maximum ORC efficiency, Several efforts resulted in systems that increase the IC engine’s
Quoilin et al. also consider the system size and cost in a paper efficiency; however, decreasing fuel prices in the 1980’s and
published in the same year [77]. These criteria are used in improved engine designs stalled implementation of the WHR
a thermo-economic optimization using six working fluids: R245fa, architecture.
R123, n-butane, n-pentane, R1234yf, and Solkatherm. Economic Recent published works sponsored by Honda, Toyota, Wartsila,
considerations find a different optimal design than a purely ther- BMW, and Volvo show a renewed interest in WHR in order to
modynamic analysis. This is due in part to the high efficiencies that increase the thermal efficiency of ICE’s [1,3,47,53,67,71]. Motivating
result from the use of high critical temperature working fluids, this resurgence is escalating fuel prices and future CO2 emission
which improve performance at the cost of increased component regulations. In addition, these systems are becoming more viable
sizes required for low specific vapor densities. The economic due to recent technological advancements that have increased the
optimum working fluid in their study is n-butane at the specific efficiencies of individual components within WHR systems. Among
cost of 2136V/kW that results in an efficiency of 4.47%, below the these improvements are smaller expander internal leakage areas
thermodynamic maximum of 5.22%. [66] and improved rotor profiles [57], along with more efficient
A valuable review of Rankine cycle thermal Exhaust Heat heat exchangers [1].
Recovery (EHR) is authored by Wang et al. in 2011 [78]. EHR is The historical review of the last section indicates that the
a sub-category of WHR limited to recovering heat at temperature majority of WHR researchers prefer steam Rankine systems or
levels provided by ICE exhaust, similar to the range of interest in ORC’s instead of Brayton, Stirling, Kalina, and supercritical CO2
this review. This work illustrates the diversity of research efforts in thermodynamic cycles along with being favored over thermoelec-
this area, along with the potential of EHR systems to improve tric devices. This is because Rankine cycle systems offer the most
overall powertrain efficiency. In comparison to Wang et al.’s effort, attractive combination of simplicity, component cost, and effi-
this paper provides historical context, practical cycle design ciency. For each intended application, the additional expenditure
considerations along with system and modeling descriptions while and complexity associated with incorporation of preheating with
highlighting additional literature in this area. As a result, the engine coolant waste heat or a recuperator (e.g. additional hard-
combination of this effort with Wang et al.’s review provides ware) above the traditional Rankine cycle should be weighed
a nearly complete summary of research in this important area. against the resulting efficiency gains. No configuration is optimal
While the authors provide insight on working fluid and for every waste heat source; hence, a thermodynamic analysis
expander selection considerations, the length of the work prevents targeting the specific source must be conducted first.
a detailed description of each study. Moreover, their efforts do not A review of the literature demonstrates that selection of
draw any definitive conclusions, as this manuscript will accomplish the working fluid and expander has a significant influence on the
in the following section. However, briefly summarized here, their efficiency of the WHR system. Most applications achieve the
paper states that reciprocating expanders are preferred for loco- highest ORC efficiencies using nearly isentropic and high critical
motives, ships, and stationary engines, despite having a lower level temperature working fluids. However, these criteria fail to address
of technical maturity. In addition, applications featuring conversion numerous practical design conditions, such as operating pressures,
of mechanical to electrical energy favor turbines, while the rota- component sizes, expander rotational speeds, expansion ratios, and
tional speeds of reciprocating expanders allow direct use of environmental concerns. Thus, the space available onboard mobile
mechanical work. Working fluid selection is crucial for efficient EHR waste heat sources should be determined prior to cycle design. In
and achievement of high cycle efficiencies. Water’s high stability addition, further costs incurred from utilizing exceedingly high or
720 C. Sprouse III, C. Depcik / Applied Thermal Engineering 51 (2013) 711e722

low pressures at specific states of the cycle to promote efficiency Since these models often neglect heat transfer to the environ-
must be justified. For instance, greater thermal efficiencies are ment along with mechanical losses, actual WHR systems rarely
achievable through higher evaporation pressures; however, this achieve the performance suggested by idealized computer models.
adds expense to the system in order to handle these high pressures As a result, physical experiments often produce roughly half the
[79]. power of the idealized WHR. On average, optimistic simulations
Environmental concerns negate the use of many of the fluids forecast ORC efficiencies between 15 and 20%, while realistic
discussed in early works. As a result, ORC systems should only expectations lie in the range of 7e10%. Thus, since engine effi-
utilize working fluids with low Global Warming Potential (GWP) ciencies are similar to exhaust waste heat fractions, fuel economy
and Ozone Depletion Potential (ODP) [76]. Natural refrigerants or improvements around 10% are a feasible outcome.
hydrocarbons are preferred, while HCFC’s and HFC’s provide fewer Direct predictions of fuel economy improvements must account
negative effects than CFC’s [76]. Organic fluids and water lead to for a slightly reduced engine performance caused by the additional
similar efficiencies in the range of normal engine exhaust exhaust backpressure imposed. In particular, horsepower losses of
temperature levels; however, organic fluids are favored at low 1.5e2.5% at high loads and 0.2e0.5% at low loads have been found
temperatures and water is preferred as the source temperature [4]. In addition, the powertrain efficiency gain only begins after
increases. While water often yields a more efficient WHR process at sufficient heat is transferred to the working fluid, which requires
these higher temperatures, condensing pressures below atmo- approximately three to four min in the presence of an accumulator
spheric pressure and wet expander exit conditions are common [63]. Systems without accumulators achieve reduced heating
deterrents to their usage. In addition, water requires high expan- times, providing a larger reduction of emissions and fuel
sion ratios that may not be met with reciprocating expanders or consumption under cold start driving cycle assessments. Most
single-stage turbines and freezing occurs with low ambient efforts locate waste heat recovery equipment downstream of
temperatures. catalytic devices. In this configuration, carbon dioxide emissions
Previous works have shown that turbines are superior for high drop in proportion to fuel consumption, with marginal effects on
temperature and high output WHR systems, while reciprocating emissions during engine warm up. Alternatively, combined evap-
expanders are preferred for small applications. Turbines generally orator/catalyst devices and upstream evaporators suffer from
provide higher efficiencies than reciprocating devices, but the prolonged startup emissions, lessening the emissions advantage of
aforementioned technological advancements have narrowed this ORC WHR systems.
disparity. In addition, turbines do have a higher initial cost, which is Research suggests a properly designed ORC for WHR of engine
offset slightly by additional fuel savings through this increased exhaust can reduce emissions and pay for itself through fuel savings
efficiency. Of interest, practical advantages of reciprocating within a relatively short amount of time (approximately 2e5 years),
expanders such as lower rotational speeds, potential hermetic primarily depending on annual travel. Hence, the system can help
power generation, and decreased sensitivity to wet expander outlet achieve future emission standards while reducing operating costs.
conditions should be considered when choosing an option for the Construction of a model that couples a virtual engine to a WHR
expansion process. Reciprocating expanders are also more system allows quantification of these advantages, while simulta-
controllable with respect to fluctuating waste heat conditions. Early neously meeting any packaging constraints or initial investment
efforts couple system expanders to output shafts with speed restrictions. To achieve the maximum ORC performance, parametric
reduction gearboxes, while later work targets hybrid vehicles as to studies of operating variables (pump speed, expander load, etc.)
decouple energy recovery from usage. Belt systems link some should be conducted to yield the highest power. At the same time,
reciprocating expanders to output shafts; however, turbine rota- second law analysis reveals where the largest amounts of exergy
tional speeds prevent this configuration. destruction are occurring. Going further, the ORC can be optimized
A large portion of potential engine WHR applications are mobile from an economic perspective, determining the configuration that
vehicles, such as passenger cars and long-haul trucks. Since the gives the highest fuel savings over a specified period of time [77].
space available for the ORC system is limited, low volumetric flow
fluids and compact heat exchangers are desirable. It is possible to 3. Conclusion
adjust for small heat exchanger size using forced convection, which
can be achieved by exposing the condenser to ram-air or using A review of the open literature with respect to WHR finds that IC
a fan-equipped condenser. However, component size may be engine exhaust contains heat with sufficient exergy to justify
limited by selecting a working fluid with desirable heat transfer implementation of a secondary cycle. Among the different cycles,
characteristics, as described by Lee [80]. With respect to pumping ORC’s are the most commonly implemented WHR systems in order
the working fluid throughout the cycle, the high fluid temperatures to generate power from low-grade to medium-grade waste heat.
achieved tend to favor the use of diaphragm-type pumps. This is This is because Rankine systems use standard components and
because they are designed to operate at high pressures and low operate efficiently using these grades of energy. Turbomachines are
flow rates, typically encountered for high ORC efficiency designs in preferred for large output systems while displacement-type
small applications such as ICE WHR. expanders dominate small-scale efforts. Finally, no single working
The largest portions of waste heat from internal combustion fluid is best for all ORC’s as the selection of the proper working fluid
engines are lost through the high- and medium-temperature requires consideration of operating conditions, environmental
engine exhaust and coolant, respectively. Studies show that the concerns and economic factors.
higher exergy level of engine exhaust allows more efficient heat
recovery; however, preheating a WHR system’s working fluid using Acknowledgements
engine coolant energy can also be desirable [51,56]. Selection of the
optimal working fluid for a specific application requires WHR This research was partially funded by the University of Kansas,
system models that are capable of comparing cycle performance Transportation Research Institute from Grant # DT0S59-06-G-
using different fluids. The fluids with the highest critical temper- 00047, provided by the US Department of TransportationeResearch
ature often appear the most desirable from a first law point of view, and Innovative Technology Administration. The sponsor supported
but practical and economic considerations should also receive the research but was not involved in the decision to submit this
attention. article.
 C. Sprouse III, C. Depcik / Applied Thermal Engineering 51 (2013) 711e722 721

References [31] D.M. Rowe, Thermoelectrics, an environmentally-friendly source of electrical

power, Renew. Energy 16 (1999) 1251e1256.
[32] Y.S.H. Najjar, Gas turbine cogeneration systems: a review of some novel
[1] T. Endo, S. Kawajiri, Y. Kojima, K. Takahashi, T. Baba, S. Ibaraki, T. Takahashi, et
cycles, Appl. Therm. Eng. 20 (2000) 179e197.
al., Study on Maximizing Exergy in Automotive Engines. SAE Int. Publication
[33] S. Quoilin, V. Lemort, Technological and economic survey of organic Rankine
2007-01-0257 (2007).
cycle systems, in: Proc. of the 5th European Conf: Economics and Manage-
[2] H. Teng, G. Regner, C. Cowland, Waste Heat Recovery of Heavy-Duty Diesel
ment of Energy in Industry, Algarve, Port, Apr 14e17, 2009.
Engines by Organic Rankine Cycle Part I: Hybrid Energy System of Diesel and
[34] F. Yang, X. Yuan, L. Guiping, Waste heat recovery using heat pipe heat
Rankine Engines. SAE Int. Publication 2007-01-0537 (2007).
exchanger for heating automobile using exhaust gas, Appl. Therm. Eng. 23
[3] M. Kadota, K. Yamamoto, Advanced Transient Simulation on Hybrid Vehicle
(2003) 367e372.
Using Rankine Cycle System. SAE Int. Publication 2008-01-0310 (2008).
[35] D. Morgan, P. Patel, E. Doyle, R. Raymond, R. Sakhuja, K. Barber, Laboratory
[4] R. El Chammas, D. Clodic, Combined Cycle for Hybrid Vehicles. SAE Int.
Test Results Low Emission Rankine-Cycle Engine with Organic-Based Working
Publication 2005-01-1171 (2005).
Fluid and Reciprocating Expander for Automobiles. SAE Publication 739062
[5] O. Badr, S. Naik, P.W. O’Callaghan, S.D. Probert, Expansion machine for a low
(1973).
power-output steam Rankine-cycle engine, Appl. Energy 39 (1991) 93e116.
[36] P. Patel, E. Doyle, Compounding the Truck Diesel Engine with an Organic
[6] M.J. Moran, H.N. Shapiro, Fundamentals of Engineering Thermodynamics,
Rankine-Cycle System. SAE Publication 760343 (1976).
sixth ed., John Wiley & Sons, Danvers, 2008.
[37] E. Doyle, L. DiNanno, S. Kramer, Installation of a Diesel-Organic Rankine
[7] C. Wu, Thermodynamic Cycles: Computer-Aided Design and Optimization,
Compound Engine in a Class 8 Truck for a Single-Vehicle Test. SAE Publication
Marcel Dekker, New York, 2004.
790646 (1979).
[8] I. Kubo, Technical and Economic Study of Stirling and Rankine Cycle Bot-
[38] M. Hall, Cogeneration: the promise is now, Mech. Eng. 103 (1981) 22e23.
toming Systems for Heavy Duty Truck Diesel Engines. Cleveland (OH): NASA/
[39] J.B. Heywood, Automotive engines and fuels: a review of future options, Prog.
DOE (1987).
Energy Combust. Sci. 7 (1981) 155e184.
[9] Y.M. Park, R.E. Sonntag, A preliminary study of the Kalina power cycle in
[40] T.J. Marciniak, J.L. Krazinski, J.C. Bratis, H.M. Bushby, E.H. Buyco, Comparison of
connection with a combined cycle system, Int. J. Energy Res. 14 (1990) 153e162.
[10] A.I. Kalina, Combined-cycle system with novel bottoming cycle, J. Eng. Gas Rankine-Cycle Power Systems: Effects of Seven Working Fluids, Argonne
National Laboratory, Energy and Environmental Systems Division, Argonne
Turbines Power 106 (1984) 737e742.
(IL), 1981 Jun, Report No. ANL/CNSV-TMe87 DE82 005599. Sponsored by the
[11] C.H. Marston, Parametric analysis of the Kalina cycle, J. Eng. Gas Turbines
Department of Energy.
Power 112 (1990) 107e116.
[41] G. Angelino, M. Gaia, E. Macchi, A review of Italian activity in the field of
[12] U. Desideri, E. Olsson, S.S. Stecco, G. Svedberg, The ammoniaewater mixture
organic Rankine cycles, VDI-Berichte 539 (1984) 465e482.
Rankine cycle: considerations on its applicability as bottoming cycle, in: Proc.
[42] O. Badr, P.W. O’Callaghan, S.D. Probert, Thermodynamic and thermophysical
26th IECEC, vol. 2, 1991, pp. 449e454.
properties of organic working fluids for Rankine-cycle engines, Appl. Energy
[13] C.H. Marston, M.J. Moran, Development of the adjustable proportion fluid
19 (1985) 1e40.
mixture cycle, Mech. Eng. 114 (1992) 76e81.
[43] J.J. Martin, Correlations and equations used in calculating the thermodynamic
[14] JC. Bass, NB. Elsner, FA. Leavitt, Performance of the 1 kW thermoelectric
properties of freon refrigerants, Thermodyn. Transp. Prop. Gases, Liquids,
generator for diesel engines, Proc. 30th Int. Conf. Thermoelectrics, Kansas City,
Solids (1959) 110.
Missouri, USA, Aug 30eSep 1, 1995.
[44] R.C. Downing, Refrigerant equations, ASHRAE Trans. 80 (1974) 158e169.
[15] E.F. Thacher, B.T. Helenbrook, M.A. Karri, C.J. RIchter, Testing of an automobile
[45] O. Badr, S.D. Probert, P.W. O’Callaghan, Selecting a working fluid for
exhaust thermoelectric generator in a light truck, Proc. Inst. Mech. Eng. Part D:
a Rankine-cycle engine, Appl. Energy 21 (1985) 1e42.
J. Auto Eng. 221 (2007) 95e107.
[46] Feasibility Test of Compounding the Internal Combustion Engine for Auto-
[16] E.W. Miller, T.J. Hendricks, R.B. Peterson, Modeling energy recovery using
motive VehicleseTask II, Thermo Electron Corporation, Waltham (MA), 1976
thermoelectric conversion integrated with an organic Rankine bottoming
(Sponsored by the Office of Highway Vehicle Systems).
cycle, J. Electron. Mater. 38 (2009) 1206e1213.
[17] H. Chen, D.Y. Goswami, M.M. Rahman, E.K. Stefanakos, Energetic and exergetic [47] H. Oomori, S. Ogino, Waste Heat Recovery of Passenger Car Using a Combi-
analysis of CO2- and R32-based transcritical Rankine cycles for low-grade heat nation of Rankine Bottoming Cycle and Evaporative Engine Cooling System.
conversion, Appl. Energy 88 (2011) 2802e2808. SAE Publication 930880 (1993).
[18] Y.J. Baik, M. Kim, K. Chang Chang, S. Jin Kim, Power-based performance [48] J. Larjola, Electricity from industrial waste heat using high-speed organic
comparison between carbon dioxide and R125 transcritical cycles for a low- Rankine cycle (ORC), Int. J. Prod. Econ. 41 (1995) 227e235.
grade heat source, Appl. Energy 88 (2011) 892e898. [49] T.C. Hung, T.Y. Shai, S.K. Wang, A review of organic Rankine cycles (ORCs) for
[19] E. Cayer, N. Galanis, M. Desilets, H. Mesreddine, P. Roy, Analysis of a carbon the recovery of low-grade waste heat, Energy 22 (1997) 661e667.
dioxide transcritical power cycle using a low temperature source, Appl. [50] T.C. Hung, Waste heat recovery of organic Rankine cycle using dry fluids,
Energy 86 (2009) 1055e1063. Energy Convers. Manag. 42 (2001) 539e553.
[20] H.E. Khalifa, Waste Heat Recovery from Adiabatic Diesel Engines by Exhaust- [51] M. Kane, D. Larrain, D. Favrat, Y. Allani, Small hybrid solar power system,
Driven Brayton Cycles, United Technologies Research Center, East Hartford Energy 28 (2003) 1427e1443.
(CT), 1983 Dec, Report No. NASA CR-168257. Sponsored by the Department of [52] B.-T. Liu, K.-H. Chien, C.-C. Wang, Effect of working fluids on organic Rankine
Energy. cycle for waste heat recovery, Energy 29 (2004) 1207e1217.
[21] M.M. Bailey, Comparative Evaluation of Three Alternative Power Cycles for [53] H. Schmid, Less emissions through waste heat recovery, Presented at the
Waste Heat Recovery from the Exhaust of Adiabatic Diesel Engines, Lewis Green Ship Technology Conf., London, Apr 28e29, 2004.
Research Center, Cleveland (OH), 1985 Jul, Report No. NASA TM-86953. [54] Wartsila, High Efficiency Heat Recovery in Service [Internet], [cited 2010 Dec
Sponsored by the Department of Energy. 2]. Available from: http://www.wartsila.com/Wartsila/global/docs/en/about_
[22] M. Bianchi, A. De Pascale, Bottoming cycles for electric energy generation: us/twentyfour7/2_2006/High_efficiency_heat_recovery_in_service.pdf.
parametric investigation of available innovative solutions for the exploi- [55] R. Stobart, R. Weerasinghe, Heat Recovery and Bottoming Cycles for SI and CI
tation of low and medium temperature heat sources, Appl. Energy 88 EngineseA Perspective. SAE Int. Publication 2006-01-0662 (2006).
(2011) 1500e1509. [56] D.A. Arias, T.A. Shedd, R.K. Jester, Theoretical Analysis of Waste Heat Recovery
[23] F.A. DiBella, L.R. DiNanno, M.D. Koplow, Labratory and On-highway Testing of from an Internal Combustion Engine in a Hybrid Vehicle. SAE Int Publication
Diesel Organic Rankine Compound Long-Haul Vehicle Engine. SAE Publication 2006-01-1605 (2006).
830122 (1983). [57] H. Leibowitz, IK. Smith, N. Stosic, Cost Effective Small Scale ORC Systems for
[24] E. Poulin, R. Demler, I. Krepchin, D. Walker, Steam Bottoming Cycle for an Power Recovery from Low Grade Heat Sources, Proc. of IMECE 2006, Chicago,
Adiabatic Diesel Engine, Foster-Miller Inc., Waltham (MA), 1984 Mar, Report Illinois, USA, Nov 5e10, 2006.
No. NASA CR-168255. Sponsored by the Department of Energy. [58] P. Colonna, H. van Putten, Dynamic modeling of steam power cycles: part Ie
[25] SS. Stecco, U. Desideri, A thermodynamic analysis of the Kalina cycles: modeling paradigm and validation, Appl. Therm. Eng. 27 (2007) 467e480.
comparisons, problems and perspectives, in: 34th ASME Int. Gas Turbines and [59] H. van Putten, P. Colonna, Dynamic modeling of steam power cycles: part IIe
Aeroengine Congress and Exposition, Toronto, Ontario, Canada, Jun 4e8, 1989. simulation of a small simple Rankine cycle system, Appl. Therm. Eng. 27
[26] M. Jonsson, J. Yan, Ammoniaewater bottoming cycles: a comparison between (2007) 2566e2582.
gas engines and gas diesel engines as prime movers, Energy 26 (2001) 31e44. [60] D. Wei, X. Lu, Z. Lu, J. Gu, Performance analysis and optimization of organic
[27] M. Jonsson, Advanced power cycles with mixtures as the working fluid Rankine cycle (ORC) for waste heat recovery, Energy Convers. Manag. 48
[dissertation], Royal Institute of Technology; 2003. (2007) 1113e1119.
[28] J. Wang, Y. Dai, L. Gao, Exergy analyses and parametric optimizations for [61] H. Teng, G. Regner, C. Cowland, Waste Heat Recovery of Heavy-Duty Diesel
different cogeneration power plants in cement industry, Appl. Energy 86 Engines by Organic Rankine Cycle Part II: Working Fluids for WHR-ORC. SAE
(2009) 941e948. Int. Publication 2007-01-0543 (2007).
[29] Y. Chen, P. Lundqvist, A. Johansson, P. Platell, A comparative study of the [62] P.J. Mago, L.M. Chamra, C. Somayaji, Performance analysis of different working
carbon dioxide transcritical power cycle compared with an organic Rankine fluids for use in Rankine cycles, IMechE Part A J. Power Energy 221 (2007)
cycle with R123 as working fluid in waste heat recovery, Appl. Therm Eng. 26 255e264.
(2006) 2142e2147. [63] R. Stobart, S. Hounsham, R. Weerasinghe, The Controllability of Vapor Based
[30] S.B. Riffat, X. Ma, Thermoelectrics: a review of present and potential appli- Thermal Recovery Systems in Vehicles. SAE Int. Publication 2007-01-0270
cations, Appl. Therm. Eng. 23 (2003) 913e935. (2007).
722 C. Sprouse III, C. Depcik / Applied Thermal Engineering 51 (2013) 711e722

[64] V. Lemort, C. Cuevas, J. Lebrun, IV. Teodore, S. Quoilin, Development and [72] I. Vaja, A. Gambarotta, Internal combustion engine (ICE) bottoming with
experimental validation of an organic Rankine cycle model, Proc. of the Heat organic rankine cycles (ORCs), Energy 35 (2010) 1084e1093.
Transfer in Components and Systems for Sustainable Energy Technologies [73] N. Yamada, M.N.A. Mohamad, Efficiency of hydrogen internal combustion
Conf., Chanbery, France, Apr 18e20, 2007. engine combined with open steam Rankine cycle recovering water and waste
[65] V. Lemort, IV. Teodorese, J. Lebrun, Experimental study of the integration of heat, Int. J. Hydrogen Energy 35 (2010) 1430e1442.
a scroll expander into a heat recovery Rankine cycle, in: Proceedings of the [74] S. Quoilin, V. Lemort, J. Lebrun, Experimental study and modeling of an organic
18th International Compressor Engineering Conference, West Lafayette, Rankine cycle using scroll expander, Appl. Energy 87 (2010) 1260e1268.
Indiana, USA, Jul 17e20, 2006. [75] W.M.S.R. Weerasinghe, R.K. Stobart, S.M. Hounsham, Thermal efficiency
[66] S. Quoilin, Experimental study and modeling of a low temperature Rankine improvement in high output diesel engines a comparison of a Rankine cycle
cycle for small scale cogeneration [dissertation], University of Liege, 2007. with turbo-compounding, Appl. Therm Eng. 30 (2010) 2253e2256.
[67] J. Ringler, M. Seifert, V. Guyotot, W. Hubner, Rankine Cycle for Waste Heat [76] J.P. Roy, M.K. Mishra, A. Misra, Performance analysis of an organic Rankine
Recovery of IC Engines. SAE Int. Publication 2009-01-0174 (2009). cycle with superheating under different heat source temperature conditions,
[68] R. Chacartegui, D. Sanchez, J.M. Munoz, T. Sanchez, Alternative ORC bottoming Appl. Energy 88 (2011) 2995e3004.
cycles for combined cycle power plants, Appl. Energy 86 (2009) 2162e2170. [77] S. Quoilin, S. Declaye, B. Tchanche, V. Lemort, Thermo-economic optimization
[69] K.K. Srinivasan, P.J. Mago, S.R. Krishnan, Analysis of exhaust waste heat of waste heat recovery organic Rankine cycles, Appl. Therm Eng. 31 (2011)
recovery from a dual fuel low temperature combustion engine using an 2885e2893.
organic Rankine cycle, Energy 35 (2010) 2387e2399. [78] T. Wang, Y. Zhang, Z. Peng, G. Shu, A review of researches on thermal exhaust heat
[70] M. Mohd.Tahir, N. Yamada, T. Hoshino, Efficiency of compact organic Rankine recovery with Rankine cycle, Renew. Sustain. Energy Rev. 15 (2011) 2862e2871.
cycle system with rotary-vane-type expander for low-temperature waste heat [79] W. Gu, Y. Weng, Y. Wang, B. Zheng, Theoretical and experimental investiga-
recovery, Int. J. Environ. Sci. Eng. 2 (2010) 11e16. tion of an organic Rankine cycle for a waste heat recovery system, Proc. Inst.
[71] N. Espinosa, L. Tilman, V. Lemort, S. Quoilin, B. Lombard, Rankine cycle for Mech. Eng. Part A J. Power Energy 223 (2009) 523e533.
waste heat recovery on commercial trucks: approach, constraints and [80] M.J. Lee, D.L. Tien, C.T. Shao, Thermophysical capability of ozone-safe working
modelling, Presented at the Diesel International Conference and Exhibition, fluids for an organic Rankine cycle system, Heat Recover. Sys. CHP 13 (1993)
France, 2010. 409e418.